Controlled Synthesis of Potential Matrix Materials and Reactive Additives
This chapter applies the MSE systems methodology to the synthesis of nonconventional concrete matrices. Specifically, this chapter discusses possible means of controlling the synthesis of conventional concrete in order to change its structure, composition, and chemistry to produce nonconventional concrete with superior performance and properties.
The synthesis of conventional Portland cement is basically the reaction of a basic oxide (CaO) with an acidic monomer (SiO2) to form a polymer. Although there are numerous other methods for synthesizing a cementitious material, very few are sufficiently low-cost or widely available for use in roads and bridges. Novel cement materials therefore must be based on one or a combination of the following strategies: (1) use of the next least expensive basic oxides other than CaO to produce the initial acidic monomer (e.g., Na2O); (2) use of the basic calcium oxide and silica in conventional Portland cement in conjunction with other materials that contain a substantial amount of acidic monomer to produce different reactions and products; or (3) use of the basic calcium oxide and silica in Portland cement but refining and controlling the reaction. Some possible combinations of these three strategies to produce a nonconventional matrix and potentially resolve some of the shortcomings of conventional concrete matrices are:
An energy efficient and environmentally benign strategy to produce matrices that incorporate waste materials (e.g., ground glass, blast-furnace slag, fly ash, silica fume) in a manner that permits the energy previously invested in producing these materials to be reclaimed.
A strategy that refines and controls the reaction to produce a less alkaline matrix that would not readily react with carbon dioxide. Other means of passivating the rebar would have to be found, however, or the matrix would have to be sufficiently impermeable to make passivation of the steel unnecessary. Although attempts to commercialize a new concrete on the basis of this strategy were previously unsuccessful (Henrichsen, 1996), new opportunities may exist that could be exploited.
A strategy that used sodium oxides to decompose and activate the aluminosilicate materials to form the matrix, thereby reducing the ion exchange problem that occurs between calcium silicate and many sodium compounds. A sodium aluminosilicate gel phase would have to be formed that would resist aqueous attack as well as the calcium silicate phase does.
A strategy that moderated or controlled the thermal stresses resulting from hydration reactions to reduce the crack density of the matrix and thereby prevent the transport of degradation factors.
Methods for controlling the synthesis of conventional concrete in order to change its structure, composition, and chemistry, and produce nonconventional concretes with superior performance and properties include the addition of agents and reactive inorganic additives to control gelation and rheology, water distribution, concrete shrinkage, temperature fluctuation, and rebar corrosion.
GELATION AND RHEOLOGY CONTROL
Improved methods for controlling the setting process of conventional concrete are essential to producing nonconventional concretes with superior properties and performance.
Theoretical and experimental advances over the past 10 years have shown that virtually all gels fall into the same universality class, which means that the fundamental physics of these gels is independent of the length scale of the gel's internal structure. The concept of universality allows the microscopic behavior of molecules and colloids to be extended to macroscopic systems, such as the sand and stone aggregates. Conventional concrete exhibits rheological behavior similar to gels on both the macroscopic and microscopic scale. The term gel, as used in this section, is a subset of the many materials referred
to as gels. In the world of physics, all systems that exhibit the rheological characteristics shown in Figure 1-16 fall into the universality class of gels.
Universality Class: A Definition
Universality class is a concept that derives from scaling behavior in the theory of critical phenomena. A percolation threshold is a connectivity critical point at which the distinction between a solid and a liquid vanishes, just as the critical point of water is the point at which the distinction between liquid water and water vapor vanishes. In the case of connectivity, many random structures, such as those of percolation clusters and random walk polymers, exhibit a fractal structure. The mass of an object generally scales as the dimensions raised to some power (e.g., for a sphere, a Euclidean object, the mass scales as the radius to the third power). The mass of a fractal object scales as its dimensions to some power less than that of space in which it is embedded. This fractal relationship means that the structures are described by “dilational symmetry” (i.e., they look the same after doubling or tripling their size). All critical phenomena fall into one or another “universality class,” which means that all objects exhibit the same scaling behavior near the critical point, because no matter how big or small, or what an object is made of, all properties that relate to the critical transition will be the same.
These fractal geometry and critical phenomena represent the most important and rigorous mathematical description of random systems that has been developed to date. Slumping is a perfect example. The same principles apply, whether applied to the collapse of a sand pile, the rheology of baking dough, or the slump test of concrete. The length scale in a sandcastle is about the size of a grain of sand and in dough is roughly a particle of flour. A single grain of sand, of course, does not slump as a result of added water, only the pile does. If one were to observe the pile of sand with a magnifying glass, one would not see fractal behavior. Concrete may exhibit fractal behavior over a huge range of length scales because one would see a range of particle sizes with a magnifying glass, in a teaspoonful, a shovelful, and even a truckload of concrete.
The most important consequence of gelation is the sol-gel transition. On the microscopic level, this transition occurs when a random growth process on either a molecular or colloidal scale generates a sol of clusters with a fractal structure. One of the properties of fractal clusters is that their density decreases as their mass increases. In principle, a cluster can become arbitrarily large with an arbitrarily small
mass. The sol-gel transition occurs when these fractal clusters get sufficiently large to start to impinge on each other and reach a percolation threshold. The material becomes a solid at this point in the sense that it exhibits elastic, rather than viscous, mechanical properties because the linked fractals are theoretically nondeformable. It is a solid with no strength, however, because there is a relatively small amount of chemical bonding between the solids in the structure. Just after the sol-gel transition, the material is exceedingly fragile. Applied stresses cannot be relieved by viscous flow, and the solid behaves as a brittle material with a very low fracture-toughness. Hence, it cracks easily.
Although this description of the sol-gel transition is based on colloidal gels, the principle of universality and the fact that the mechanical properties of a concrete body are the same as for the colloidal gel mean that a conventional concrete structure will be transformed from a liquid to a very weak, brittle solid at some point. This state in the gelation process (the “set” in concrete terminology, as described in Chapter 1) has two major consequences for the material.
The material is susceptible to a large amount of damage during placement. As a practical matter, the rheology of the concrete as it is being poured and the gelation and setting during curing are intertwined. Under the large shear stresses associated with pouring and forming, there is no real distinction between a gel that is easily fractured into small pieces and a viscous liquid containing aggregate.
The material is easily subject to cracking from vibration or unintentional loading in its first few hours that can affect its durability. Although the cracks may be sufficiently small that the aggregate in the concrete can bridge them and prevent significant degradation of the bulk mechanical properties of the structure, the cracks permit the ingress of water, salt, and CO2, making it more susceptible to chemical degradation.
There are three possible methods to control the synthesis of the concrete matrix to produce a nonconventional concrete that has superior properties and performance:
Gelation acceleration. If gelation could be accelerated after the material has been poured, the opportunity for further damage could be reduced. This acceleration might be achieved by
microwave drying, for example, which has been common in the ceramics industry for some time and more recently been applied to accelerate the setting of cement. Wu et al. (1987) report that 15 to 30 minutes of microwave treatment enhanced the strength and lowered the permeability of the hardened concrete. Hutchison et al. (1991) showed that the induction period for the hydration of cement was reduced by microwave heating. More recently, Johnson (1996) has attempted to quantify such effects.
Gelation inhibition. If the sol-gel transition of the matrix is delayed, the concrete will remain workable for a longer time with less water, making it less susceptible to cracking during placement. A cement that did not gel like conventional cement would likely have much better and more reproducible properties. It is well known in the laboratory that surfactants retard or suppress gelation. It is no coincidence that the current additives called “high-range water reducers,” which can make concrete workable at much lower water-to-cement ratios, are surfactants. Surfactants work by preventing either the formation of fractal clusters or their percolation. In addition to surfactants, there are other possible strategies that could be explored for inhibiting gelation. Sols also gel as a result of changing their surface charge, by changing the pH, by suppressing the effect of repulsive surface charges through the addition of an electrolyte that screens the charge, or through the addition of another sol with an opposite surface charge. One of these approaches could be used to inhibit the gelation of nonconventional concrete.
Gelation elimination. Although gelation is an important stage in the synthesis of conventional Portland cement, it is by no means obvious that gelation is essential or desirable in a concrete matrix. A potentially rich field of innovative study would be the elimination of gelation and the development of methods in which the matrix would solidify by becoming progressively more viscous, as occurs in glass. Changing the reaction pathway to eliminate gelation may require the use of higher temperatures in the setting process or a significant departure from CaO-SiO2-Al2O3 chemistry.
AGENTS TO CONTROL WATER AND SHRINKAGE
As stated in Chapter 1, hydrated cement has a smaller volume than the sum of the volumes of the dry cement and the water that becomes
chemically bound. This contraction of approximately 8 percent can cause the development of microcracks. The contraction is further aggravated by loss of water through evaporation, causing cracking at the surface of the structure to be even more severe. Potential nonconventional methods for controlling the release of water to maintain more stable conditions during the hydration reaction could help prevent microcracking. Potential methods for water and shrinkage control include the addition of synthetic polymers and cellulose derivatives, wood and paper waste materials, swelling clays, and high-range water reducers. Research into commercial production of shrinkage control agents is currently under way (Henrichsen, 1996; N.S. Berke, personal communication).
Synthetic Polymers and Cellulose Derivatives
Certain synthetic polymers and cellulose derivatives have the ability to absorb large amounts of water and have been used in oilwell cements. Such materials might be of use by releasing water in a controlled manner and thus regulating the temperature cycle and maintaining more constant hydration-reaction conditions, albeit at the expense of increasing the amount of water required for effective processing. They could also prevent water from draining out of forms and have a beneficial effect on the rheology and workability of concrete. Additionally, fibers used for these purposes might increase the strength of the cement during the sol-gel transition and, because they shrink when dry, help to prevent cracking by putting the material in a mildly compressive state. This appears to be an extremely fruitful area for further research. An example of a potentially nonconventional concrete system is given below.
Considerable research has been conducted on the effects of fiber reinforcement of concrete (Wang, 1996). Virgin polypropylene in the range of 0.5 to 1.0 percent and carpet waste fiber in the range of 1 to 2 percent have been added to concrete to improve performance under tension and enhance crack prevention by reducing the friability of the concrete structure between setting and hardening. Hydrophilic synthetic polymers (e.g., poly[vinyl alcohol] [PVOH], hydroxyethyl cellulose) could also play multiple roles in the setting process by serving as water-control agents, crystal habit modifiers, and sources of organic material for
“fossilization” replacement reactions. PVOH has been shown to play a variety of positive roles in cement and concrete (Wang et al., 1987; Chu et al., 1994; Wang et al., 1994; Robertson, 1995). For steel fiber-reinforced cement bricks, the addition of 1.4 percent PVOH enhanced both bonding strength and frictional resistance. It is postulated that PVOH may coat fiber surfaces, helping to form a dense, fine-grained, ductile, and hydrophilic interfacial layer adjacent to the steel fibers. PVOH also inhibits the nucleation of calcium hydroxide crystals while possibly increasing the nucleation of calcium silicate hydrate. PVOH may also serve a lubricant function, allowing better packing of cement particles around steel fibers. It may also improve the cement–aggregate bond.
Wood and Paper Waste Materials
It may be possible to use sawdust or paper pulp as water-control agents. Shredded newsprint might also help to reduce micro-cracking by acting as a reinforcement—similarly to its role with flour paste in papier maché—and by reducing the friability of concrete between setting and hardening (Lin et al., 1994). These organic materials may also promote a chemically more-reducing environment that could serve to protect the rebar. Recycled paper pulp that is no longer suitable for paper formation is a relatively inexpensive source of organic material.
A potential risk of using organic materials is that they tend to decompose. However, if such materials as wood and waste paper materials can be made to act as source material for “fossilization” replacement reactions, this tendency could actually be beneficial. Fossilized wood results from the reaction of minerals in groundwater with the acidic decay products of cellulose and lignin in the wood. Similarly, sawdust and paper might neutralize, precipitate, and immobilize calcium and silica to create more C-S-H. Since an appreciable amount of free, saturated water occurs in cracks, the mineralization of material in the crack would help to heal it. It may be possible to pretreat sawdust or newsprint to make it more reactive. It may also be possible to post-treat cement with short-fiber pulp from paper recycling.
Clays that swell substantially, such as montmorillonite, could also act as water-control agents, possibly modified to hold large amounts
of water in the initial stages of setting and to release water as calcium ions are released into solution, thus deflocculating the clays and shrinking them. This may improve the rheological properties of cement. The clays would then react in the later stages of the setting process to become part of the matrix and ensure a good pre-structure for durability. Although these clays have proven to be unsuitable for conventional concrete, rheological studies have shown that they can counteract self desiccation and increase dispersion of silica fume, for example (Henrichsen, 1995).
High-Range Water Reducers (Superplasticizers)
Water-reducing agents that decrease the usual w/c ratio by 5 to 15 percent are called plasticizers, while those effecting a reduction approaching 30 percent are HRWRs, or superplasticizers. Such agents currently make it possible to produce high-strength concrete, particularly when used in conjunction with silica fume. If the normal w/c ratio is maintained, a “flowing” or self-leveling concrete is produced. The two principal materials used as HRWRs are sodium salts of sulfonated melamine formaldehyde (SMF) condensates and sulfonated naphthalene formaldehyde (SNF) condensates, both of which are anionic linear polymers. Lignosulfonates are also commonly employed. Timing is significant. The greatest effect is achieved if the HRWR is added to the cement a few minutes after mixing (Taylor, 1990).
Recent research, while focusing on the use of HRWRs in conventional concrete, provides an example of how the study of the roles of additives on the molecular level can lead to the production of high-durability concretes and new casting technologies, thus illustrating the benefits of the systems approach. Optical and scanning-electron microscopic examination of rapidly frozen samples suggest that water reduction is achieved by improving the dispersion of the cement grains in water, decreasing or preventing flocculation, and freeing the water normally trapped in the flocs. Factors affecting deflocculation include an increase in the zeta-potential, solid–liquid affinity, and possible steric hindrance (Taylor, 1990). Reviewing the mechanisms of HRWR action, Sakai and Daimon (1995) conclude that dispersion of cement particles is improved by electrical repulsion and steric hindrance effects. They argue that new types of HRWRs will play a significant role in the development of high-strength and high-durability
concretes and in the development of new casting technologies, such as highly flowable concrete.
THERMAL CONTROL AGENTS
The hydration of Portland cement is an exothermic process (Figure 1-17). In large pourings, the heat released can raise temperatures to levels at which unwanted side-reactions occur. Given the large thermal expansion coefficient of water, thermal gradients cause differential expansion during the exothermic setting process and thermal tensile stresses during final cooling, both of which cause stress cracking. Cracking problems are especially severe when the paste is beginning to form a gel network and, as stated above, is effectively a solid but has almost no strength. Cracks are thus easily initiated and do not flow together and heal. The diagram of phase transformation versus time in Figure 1–5 is for a single temperature and does not take into account the above effect. The superposition of temperature on the diagram would be analogous to a TTT diagram for steels.
The water-control agents discussed in the previous section may also help control thermal changes by slowing the rate of hydration and giving more time for the heat generated to dissipate without the development of significant temperature gradients. These water-control agents might also help by maintaining a more constant chemical potential of the water, thus causing the reaction rate and the heat release to be more uniform.
Other potentially innovative methods for controlling the matrix gelation process are by the use of additives that cause endothermic or exothermic side-reactions.
Components may be introduced that could produce endothermic side-reactions. Pozzolanic or supplementary cementing materials can reduce the total heat of hydration. These side-reactions may absorb heat at the same rate at which it is produced and act as a thermal buffer. They may even be able to reverse and release heat later in the reaction, thus continuing the thermal buffering action.
An alternative approach to thermal control is to introduce exothermic side-reactions, such as the oxidation of powdered iron. Such side-reactions would maintain the pour at above ambient temperature but would release heat uniformly throughout the
body. These side-reactions could even be designed to react more rapidly at the surface, canceling thermal gradients. Exothermic side-reactions could also be used to make the setting temperature more independent of climate. This reaction would be particularly advantageous in colder weather, resulting in more uniform setting, and possibly could also be used to prevent freezing in cold weather.
REACTIVE INORGANIC ADDITIVES
Many inorganic substances (e.g., fly ash, silica fume, blast-furnace slag) have been incorporated as “additives,” or supplemental cementitious materials, in cement mix designs. Some are added as inexpensive extenders, but most are reactive and combine with the Ca(OH)2 to produce more C-S-H. Although the effects of these additives have been studied by the empirical measurement of the resultant properties of the cement mix, their detailed role in affecting the microstructure is poorly understood. At best, they can actually improve the performance of the concrete (e.g., reduce permeability and refine the microstructure and pore structure) and at their worst do no harm, reduce cost, and utilize what would otherwise be industrial waste.
In a MSE systems approach, these reactive inorganic materials would be considered an integral and important part of the matrix and would be integrated into the chemical design of the cement. Three examples of how these additives would be incorporated within a MSE systems approach are:
Naturally occurring materials, such as volcanic ash and diatomaceous earth, referred to as “pozzolans,” have been used as cement components since Roman times. These materials usually have small particle size and are amorphous. Their beneficial function is to supply additional reactive silica and alumina to the cement paste. The additional silica reacts with the calcium hydroxide component of the cement, producing the more stable C-S-H phase more quickly and facilitating curing. It would also react more slowly than the silica and alumina components of the cement, delaying the sol-to-gel transition. If these materials are also rich in alumina, they can hydrate in an alkaline environment and compensate for excess water.
Unreacted materials are inherently more durable as vitreous substances than xerogels.1 By providing acidic oxides, the effective pH of the pore phase would be reduced in a controlled manner rather than by the vagaries of such environmental factors as CO2 and acid deposition.
Processes could potentially be developed that would purify and activate these inorganic reactive materials, just as silicate phases are treated with calcium carbonate and “purified” of water and carbonate to create clinker. If close regard were paid to the types of chemical reactions that were desired in the cement, these materials could be activated by an acid treatment (e.g., hot water or steam) that would generate additional reactive acidic silanol groups in their surface layers and remove soluble contaminants, such as sulfates. Other nonconventional reactive components could include colloidal silica and/or alumina or recycled container glass.
As stated in Chapter 1, one difficulty with incorporating SCMs into a concrete system is that, as waste materials, their manufacture is not controlled and their chemistries may vary. Within an MSE systems perspective, SCMs would have to be characterized before use to ensure that the proper levels of uniformity and quality are present to achieve the effects discussed above.
REBAR CORROSION CONTROL AGENTS
The ancillary role that conventional cement plays in passivating the rebar poses some significant challenges for a nonconventional cement. As stated in Chapter 1, the calcium hydroxide component that generates the high-pH conditions and prevents the rebar from corroding also tends to dissolve rather rapidly under natural water conditions. A nonconventional cement matrix that would be more stable than conventional cement would thus require new strategies for rebar passivation. There are four approaches to this problem: (1) employ some other chemical system in the matrix to passivate the rebar;
A xerogel is a gel that has dried under near-ambient conditions, in contrast to an aerogel, which is dried under supercritical conditions, or a hydrogel, which is not dried. Whatever their conditions of formation, gels have considerable surface area and porosity that vitreous materials do not.
(2) devise a matrix that prevents external aqueous solutions from coming in contact with the rebar; (3) apply a coating to the rebar placement; or (4) develop a replacement for the rebar that does not degrade in high pH environments.
The first approach would require the development of a matrix with a chemical attribute other than pH to passivate the rebar. Redox reactions could passivate the rebar in a more neutral chemical environment and lead to a far less expansive Fe+2 oxide coating. Substances such as sulfide ores or blast-furnace slags could create a sufficiently reducing environment to reduce the rate of rebar corrosion. Organic materials (e.g., sawdust, paper pulp, polymers) may also function as reducing agents and limit the rate of rebar corrosion, which is dependent on the availability of oxygen, while helping maintain low pH. Many of these substances have already been discussed as having other roles in the matrix. Clever mechanical and chemical design within an MSE systems approach could exploit this potential synergism. For example, slag can contribute both activated silicates and reducing agents or polymers could be designed to act as both water-control agents and reductants. Of course, only substances that do not themselves produce expansive corrosion products would work. It will also be important to design structures so that these sacrificial agents are not exhausted prematurely or are exhausted at the end of the structure's useful life. A truly nonconventional approach would be to develop a chloride scavenger to prevent the migration of chlorides to the surface of the rebar. Calcium aluminates to some extent act as chloride scavengers.
The second approach would be to develop a nonconventional matrix that would not allow significant amounts of water to come in contact with the rebar. Such a matrix could not have any continuous porosity and could not generate cracks unless they were self-healing. Self-healing could occur if a substance was added that was more soluble in the oxidizing environment of a large open crack than near the rebar. As solutions passed through the crack, the sealing agent would leach out of the matrix and then precipitate in the more reducing environment near the rebar. Conceivably, a “smart” corrosion inhibitor could be developed in which the mobile agent passivated the rebar at the places were it was most exposed to chlorides or, for materials with low pH, to water.
The third and fourth approaches concern changes to the actual reinforcement and thus will be discussed in the next chapter.
EVOLUTION OF STRUCTURE OF CONCRETE
The linkages among the four points of the MSE tetrahedron in Figure 1-1 (i.e., synthesis/processing, structure/composition, properties, and performance) are not understood for conventional cements and concrete but are clearly important. An understanding of these linkages is required if better performing concretes are to be produced. Just as the key to improved metals and ceramics has been the understanding and control of microstructural evolution, so must it be for cements and concrete.
Increased understanding of these linkages might make it possible to develop such new technologies as crystalline or macromolecular nucleating agents. The addition of nuclei to conventional Portland cement at key points in its synthesis might make it possible to control the incubation time of setting, improve the spatial and temporal homogeneity of the material, or crystallize the gel phase into more chemically durable phases. An analogous area of research is biomimetic synthesis.
Biomimetics is the study of biological materials to determine the structure and properties of biocomposites and the organismal strategies in their synthesis in order to apply this knowledge to the formation of novel materials with useful engineering properties. Biological composites are composed of inorganic filler phases that are highly organized and structured inside a macromolecular matrix. Examples include mammalian bones and teeth, mollusk shells, echinoderm skeletal units, and many other biocomposites in which the inorganic phases take the form of structural units with morphological, crystallographic, and geometrical order. The physical and chemical properties of biological composites are orders of magnitude better than any of their synthetic analogs.
The structural control in layered materials is influenced by an organic matrix that acts as a template in the assembly of the inorganic units and the forming of microstructures tailored for specific physical properties. The constituent species of the inorganic phases are fairly simple (e.g., CaCO3), but because the units are highly organized in two- and three-dimensional configurations, the resultant properties are highly isotropic, multifunctional, and specifically designed to
provide the necessary physical properties to the organism that produces it.
Biological hard tissues may provide important lessons for concrete research and development. The constituent phases are simple, easily accessible, and synthesized in ambient aqueous conditions. Thus, novel strategies for the design and synthesis of nonconventional concrete materials could potentially be developed from the study of biological materials. This could include structural design guidelines for multifunctional, functionally gradient, and laminated or three-dimensional micro- and macro-architectures that incorporate smart materials and self-healing concepts. Furthermore, organic macromolecules may be useful as templates in the structure of nonconventional concrete, either as nucleating agents for filler particles and fibers or as growth modifiers to control particle morphology and shape to achieve anisotropic properties.